目的：分析急性肾小管间质性肾炎（acute tubulointerstitial nephritis, ATIN）患者的临床及实验室检查特征，探讨该病与急性肾小管坏死（acute tubular necrosis nephritis, ATN）的鉴别诊断要点。方法：纳入2009年1月至2018年12月间在上海交通大学医学院附属瑞金医院肾脏科经肾活检病理诊断为ATIN、ATN的患者，收集其临床表现和实验室检查数据，并对两者进行比较分析。结果：10年间我院肾脏科行经肾活检的病例总计5 537例，其中诊断为ATIN的患者共135例（2.4%，135/5 537），诊断为ATN的患者109例（2.0%，109/5 537）。ATIN占急性肾脏病（acute kidney disease, AKD）肾活检患者的21.4%（135/630）。ATIN患者的中位确诊年龄为53岁，女性占57.0%，临床主要表现为发热、皮疹、关节痛，常见诱因为感染、药物应用和毒物接触。与ATN组相比，ATIN患者中女性占比高（57.0%比33.9%），平均体重指数（body weight index，BMI）低（22.9±3.6比24.6±3.9，P<0.01），且发生急性肾损伤（acute kidney injury, AKI）（14.8% 比64.2%）、少尿（17.0%比48.6%）及入院后需紧急透析（19.3%比39.4%）的百分比低（P<0.01）。ATIN患者入院时，血红蛋白（hemoglobin，Hb）[（100.9±20.9） g/L比（116.7±29.8） g/L]和血尿素氮/肌酐比值（blood urea nitrogen/creatinine ratio， BCR）（11.8±5.4比14.6±11.0）均较ATN组低（P<0.01）。多因素回归分析显示，入院时高白蛋白（>55 g/L）、低血清肌酐（serum creatine，Scr）（<62 μmol/L）、低血清尿酸（urine acid，UA）（<208 μmol/L）、低Hb水平（<130 g/L）与ATIN相关，联合这4项指标建立的预测模型，其诊断ATIN的受试者操作特征曲线下面积为0.798（95%CI为0.742~0.853），灵敏度为74.4%，特异度为71.4%。结论：ATIN在上海地区人群接受肾活检的AKD患者中占比高，好发于中年女性，半数患者诱因不明。患者入院时的白蛋白、Scr、Hb及UA水平有助于ATIN与AIN间的鉴别，基于上述4项指标构建的ATIN患者诊断预测模型具有较好的特异度和灵敏度。

Objective: To analyze the clinical and laboratory features of patients with acute tubulointerstitial nephritis (ATIN) and its differential diagnosis with acute tubular necrosis (ATN). Methods: From January 2009 to December 2018, patients with ATIN or ATN confirmed by renal biopsy in the Department of Nephrology, Ruijin Hospital ，Shanghai Jiao Tong Universtiy School of Medicine were included. The clinical manifestations and laboratory examination data were collected and analyzed comparatively. Results: During the period, there were 5 537 patients receiving renal biopsy,inclu-ding 135 ATIN cases [2.4% (135/5 537)] and 109 ATN [2.0%(109/5 537)].ATIN accounted for 21.4% (135/630) of the acute kidney disease (AKD) patients. The median age of ATIN patients was 53 years old and the female proportion was 57.0%. Main clinical manifestations of ATIN included fever, rash and arthralgia. Infection, and medicine use and toxic exposure were common inducements of ATIN. Compared with the ATN group, ATIN patients had higher female proportion (57.0% vs 33.9%) and lower BMI (22.9±3.6 vs 24.6±3.9, P<0.01). The occurrence of acute kidney injury (AKI) (14.8% vs 64.2%), oliguria（17.0% vs 48.6%）, and emergency dialysis after admission （19.3% vs 39.4%）in ATIN patients is lower (P<0.01). Hemoglobin (Hb) level [(100.9±20.9) g/L vs 116.7±29.8 g/L)] and blood urea nitrogen/creatinine ratio (BCR) (11.8±5.4 vs 14.6±11.0) in patients with ATIN were lower than those in ATN group(P<0.01). Multivariate regression analysis showed that high serum albumin（55 g/L）, low serum creatine (Scr) (<62 μmol/L), low uric acid (UA) (<208 μmol/L), and low Hb levels(<130 g/L) on admission were associated with ATIN. The predictive model combining the above four indicators showed that the area under the curve for diagnosing ATIN was 0.798 (95%CI: 0.742-0.853), with a sensitivity of 74.4% and a specificity of 71.4%. Conclusions: ATIN accounts for a high proportion in biopsy-confirmed AKD patients from Shanghai. It is more common in middle-aged women. Half of the patients have unknown causes, and their clinical manifestations are similar to ATN. Serum albumin, Scr, Hb, and UA levels at admission are helpful to the differential diagnosis of ATIN and AIN. The diagnostic prediction model based on the above four indicators showes good specificity and sensitivity.